The Exploration of an Air Pollution Hazard Scenario using Dispersion Modeling and a Volumetric Geographic Information System
Michael Sean Scott
Advisor: Dr. Susan L. Cutter
Air pollution as an environmental hazard is of great concern as the world approaches the 21st century. Many people are concerned about their risk from exposure to elevated levels of chemicals. It is difficult, however, to measure the magnitude of an air pollution threat over even small areas, at different altitudes, and over several time periods. Mathematical dispersion models are often used to predict chemical concentration levels. Geographic information systems (GIS) have been used successfully to visualize and analyze ground-level concentrations of dispersed pollutants. Unfortunately, current GIS lack the capability to examine a pollutant plume as it truly exists - in three-dimensional space.
Therefore, the purpose of this research was to expand the analytic capabilities of traditional GIS into the third spatial dimension and implement a mathematical dispersion model in a spatial context relating to an air pollution hazard scenario, namely the release of caprolactam from AlliedSignal Inc. in Irmo/St. Andrews, South Carolina. These goals were accomplished by extending map algebra techniques (such as the addition or subtraction of map layers) to volumes. The map algebra tools were created in a modular form, allowing them to be recombined using cartographic modeling. The Industrial Source Complex - Short Term (ISCST3) dispersion model was executed to determine the characteristics of the plume emanating from AlliedSignal in all three spatial dimensions as well as to test the model's sensitivity to variable changes.
As a result, a functional prototype volumetric GIS has been developed with 49 separate operators. By using this volumetric GIS and the ISCST3 model, the Whitehall neighborhood north-northeast of the AlliedSignal facility was found to bear the greatest pollutant burden with the majority of the released caprolactam being concentrated within the facility boundary and 10 to 30 meters above the ground. In all, the average concentrations found in nearby areas should not be enough to cause acute reactions but the effects of long term exposure are less certain. Additionally, the ISCST3 model was found to be extremely sensitive to changes in stability and less sensitive to changes in wind speed and emission rates. Inconsistent patterns resulted from the inclusion of topography.
KEYWORDS: Volumetric GIS, 3D GIS, Air Dispersion, Air Pollution, South Carolina